Picking the model is the easy part. The harder calls — where to run, which platform to bet on, and how to keep your agents portable — are what determine whether an agentic AI program actually scales.
Read the full take →Cat Wu, Anthropic's Head of PM, shared three insights on Lenny's Podcast that cut straight to why most people aren't getting real value from AI yet; and what the ones who are doing differently.
Read the full take →A finance exec I spoke with recently vibe coded his way to a credible replacement for parts of his Workday and Concur stack. The SaaS moats that held for a decade are narrowing, and vendors coasting on comfort are more exposed than they realize.
Read the full take →Organizations that abandon agentic AI projects at the first stumble are making a costly mistake. The gap between 85% and 97% accuracy is a prompting problem, not a model problem — and the companies that push through are building real advantages.
Read →AI strategy perspectives for business leaders. No fixed schedule — just substance when there's something worth saying.